Background
Spinal cord injury (SCI) is a severe clinical condition affecting more than 1.3 million people in the USA [
1]. In addition to impaired motor function, SCI leads to multiple organ dysfunction and complications and increased susceptibility to pathogen infection [
2]. SCI and other disorders of the nervous system are known to induce severe immune depression resulting in increased risk of infections [
1,
3‐
5]. Specifically, respiratory infections including Influenza A virus (IAV) and pneumonia occur frequently after SCI and are the leading cause of re-hospitalization and mortality in SCI patients [
4‐
6]. Emergence of pandemic flu strains in the past decade have heightened the awareness that immune-compromised patients, such as those suffering from SCI, are more susceptible to a new strain of influenza virus [
7]. Therefore, reducing complications from respiratory infections and understanding the mechanisms contributing to immune dysfunction are critical for improving the health and life span of SCI patients.
Intact neuroimmune communication is essential for mounting a proper immunological response to infections. Neural pathways that innervate peripheral organs regulate peripheral immunity through neural reflex circuits [
8]. For example, immune cells express receptors for neurotransmitters including acetylcholine and norepinephrine (NE) which regulate immune activation [
9]. Disrupted neuroimmune communication, such as that following SCI, results in a syndrome called CNS injury-induced immunodepression (CIDS), which is characterized by increased susceptibility to infections and worse neurological outcome [
3]. Several groups, in addition to ours, have investigated the effects of SCI on immune dysfunction in models of acute and chronic SCI. Deficits in peripheral immunity with decreased antibody production by B cells [
10,
11] and impaired cytokine expression by T cells [
12,
13] have been reported in mouse models. Previously, these deficits have been attributed to sympathetic nervous system disruption. High thoracic (T3) injury disrupts essential sympathetic regulation of lymphoid organs and leads to impaired antibody synthesis and increased splenocyte apoptosis [
10,
14]. These deficits may be related to elevated splenic norepinephrine (NE) levels since blocking NE signaling by β2 adrenergic receptor inhibition increased antibody production [
10] and decreased splenic atrophy [
14] after SCI.
In addition to altered sympathetic activity and elevated splenic NE, SCI also activates other stress responses. SCI can activate or dysregulate [
15] the hypothalamic–pituitary–adrenal (HPA) axis leading to increased corticosterone (CORT) in the blood. Importantly, CORT levels were increased following both high- and mid-thoracic injures. For example, CORT levels increased rapidly after SCI [
10,
16] and remained elevated 4 weeks after mid-thoracic SCI [
14]. Because CORT has known anti-inflammatory effects [
17], these findings suggest that CORT could interfere with immune function during acute and chronic SCI and that its effects are not limited to high thoracic injuries.
In our previous studies, chronic mid-thoracic (T9) contusion injury leads to impaired CD8 T cell function [
12,
18]. We determined that chronically injured mice infected with H3N2 (X31) influenza virus were not able to mount an effective antiviral immune response. Following intranasal challenge with X31, SCI mice had significant mortality while non-injured mice exhibited no mortality [
18]. Importantly, the increased mortality was associated with impaired generation and effector function of virus-specific CD8 T cells [
18]. Interestingly, NE levels were elevated in the spleen of injured mice at this chronic timepoint [
12]. However, mice with mid-thoracic injury also experience immune depression acutely after SCI [
13] when NE levels are unchanged [
10]. Therefore, we hypothesized that deficits in CD8 T cell function can occur independently of disrupted sympathetic regulation. To test this hypothesis, we investigated CD8 T cell antiviral responses to influenza virus after acute mid-thoracic injury. Our results show that mice with acute SCI have impaired CD8 T cell function, indicating that immune depression can occur independently of disrupted sympathetic splenic innervation and elevated NE.
Discussion
Proper communication between the nervous system and peripheral immune system is necessary to maintain immune homeostasis and mount an immunological response to infection [
8]. Peripheral neurons respond to infections through cytokine receptors and pattern recognition receptors [
25]. In return, the nervous system itself can also activate immune cells. Neurogenic inflammation arises following release of inflammatory mediators from peripheral nerve terminals and has a direct effect on peripheral immune cells. These mediators include neuropeptides, neurotransmitters, and chemokines which can activate immune cells and facilitate immune cell recruitment, providing a positive feedback loop [
9,
26]. Given the high innervation of the neuronal network, and the speed of neuronal transduction, neuroimmune communication can allow for rapid immune cell activation and mobilization [
26]. Recent studies have investigated the significance of neuronal regulation of CD8 T cell function [
18] and have identified immune dysfunction after neuronal damage, such as SCI. Following viral infection, expansion of virus-specific CD8 T cells is necessary to eliminate the infection. Therefore, restoring CD8 T cell function provides a therapeutic strategy for improving resistance to influenza virus after SCI. Understanding the mechanisms of immunological failure to respond to influenza and other infections in injured mice is highly significant and clinically relevant for injured patients who may suffer from life-threatening clinical complications from not only influenza infection but also other microbial infections.
In this study, we show that mice with acute SCI had exaggerated and prolonged weight loss following influenza virus infection. The peak of weight loss occurred at 7 days post-infection in uninjured mice, while SCI mice continued to lose weight though 8 days (Fig.
1b). In addition, several SCI mice had to be removed from the study because of severe weight loss when weight loss resulting from the injury was taken into consideration. We have previously shown that mice with chronic SCI had increased mortality following virus infection [
18]. We surmise that a higher dose of virus would be completely lethal in injured mice; however, in this study a lower dose of virus was used to limit lethality and allow for further cellular studies. Seven days after infection, SCI mice had attenuated virus-specific CD8 T cells in both the lung and spleen (Figs.
3 and
4). There were both decreased total CD8 T cells and virus-specific CD8 T cells in the lung of SCI mice (Figs.
2 and
3), suggesting decreased proliferation of resident and specific CD8 T cells. Furthermore, we hypothesize that the attenuated response in the spleen contributes to decreased infiltration of specific CD8 T cells in the lung of SCI mice after infection, as other studies have suggested that splenic CD8 T cells can contribute to viral clearance in the lung [
24]. Because neuronal signaling can be involved in immune cell recruitment [
26], this lack of infiltration can be directly due to impaired neuroimmune communication following injury. Importantly, the impaired CD8 T cell response in lymphoid tissues can have implications for infections in various organs [
13] in addition to systemic infections. These data indicate impaired CD8 T cell antiviral function acutely after SCI which leads to severe weight loss and prolonged recovery from infection.
Recent studies have begun to elucidate mechanisms of immunosuppression after SCI. For example, a recent study has implicated impaired regulation of the sympathetic-neuroendocrine adrenal reflex [
15]. In addition, SCI-induced immune dysfunction has been shown to be due, in large part, to massive reorganization of the spinal sympathetic reflex circuit, e.g., the recruitment of glutamatergic interneurons, that results in increased sensitivity of this circuit [
27]. These studies, however, only show immunosuppression following high thoracic (T3) injury. Importantly, our studies show impaired antiviral immune response following mid-thoracic (T9) injury. The mechanisms of immunosuppression following T9 injury are less understood. Here, we propose that the regulatory mechanism induced by SCI, such as increased Tregs, PD1, and CORT, in addition to an unbalanced immune response can interfere with antiviral immunity. However, more detailed mechanistic studies are required to fully understand how SCI leads to impaired antiviral immunity following T9 injury.
SCI causes massive inflammation locally and also systemically [
2,
28]. Inflammatory cells are released into the blood stream and can then infiltrate both the injured spinal cord and also secondary organs, including the lung. Here, we show infiltration of neutrophils and macrophages into the lung following SCI (Fig.
5). This is important because exaggerated neutrophil accumulation can impair CD8 T cell proliferation and activation [
29,
30] and thus potentially impair viral clearance. Furthermore, there was an unbalanced cellular response to infection in the lung of SCI mice. SCI mice had decreased accumulation of CD8 T cells while innate immune cells, including macrophages and neutrophils, were exaggerated (Fig.
2). It is unknown if this unbalance in immune cells contributes to impaired CD8 T cell function. Innate immune cells like neutrophils, macrophages, and dendritic cells are important for activating and recruiting CD8 T cells during infections. However, altered activation state and function following SCI may interfere with their normal roles during infection. Further studies are needed to determine whether chemotactic signaling by innate immune cells is impaired following SCI. The reason why neutrophils accumulate in the lung after SCI is still unknown, but may be related to leaked gut microbiota that has translocated into lung tissue as recently reported [
31]. Overall, our finding of increased neutrophils in the lung of SCI mice, both before and after infection, could have serious implications for pulmonary function [
32] and antiviral responses.
Cellular profiling showed upregulation of several immunomodulatory pathways (PD-1/PD-L1, Tregs) in the spleen of injured mice (Fig.
5). Although the mechanism of how PD1 and Tregs are induced following SCI is still unknown, it is understood that these pathways are upregulated to counteract the systemic inflammatory response in order to limit further spread of inflammation at the injury site and also protect against autoimmunity [
2]. For example, Tregs can promote tissue remodeling and depletion of Tregs during SCI worsened outcome [
33]. The role of PD1 expression on T cells during recovery from SCI has not been studied in detail; however, PD1 expression on macrophages within the injured spinal cord was important for limiting inflammation and improving recovery [
34]. Although PD1 and Tregs have roles in healing and recovery, they can also interfere with antiviral immune function during infection. We have previously shown that PD1 upregulation during chronic SCI decreases CD8 T cell activation [
12]. In addition, Tregs suppress antiviral immunity which can enhance viral replication [
35]. Overall, it is possible that the regulatory mechanisms set in motion following injury lead to attenuated responses to influenza challenge. Given the importance of the spleen as a lymphoid organ, the upregulation of these pathways could contribute to the impaired response to both systemic [
13] and respiratory infections.
Deficits in peripheral immunity could also be attributed to sympathetic nervous system disruption and enhanced splenic NE signaling. Adrenergic nerve terminals in the spleen and liver regulate NE release [
9]; however, this regulation becomes impaired following injury, leading to enhanced NE levels and decreased immune activation. Previous studies have shown that T3 transection which completely disrupts essential sympathetic regulation of lymphoid organs leads to increased splenocyte apoptosis and impaired immune function that was dependent on NE signaling [
10,
14]. In addition, we have previously shown that NE increases PD1 expression and directly decreases activation of CD8 T cells [
12]. However, NE levels are unchanged early after T9 injury [
10,
14]. In addition to neurotransmitters like NE, altered neuroimmune communication and increased neuroinflammation can also increase levels of stress-related hormones, including CORT. SCI activates or dysregulates the HPA axis [
15] leading to increased CORT in the blood. For example, CORT levels increased rapidly after SCI [
10,
16] and remained elevated 4 weeks after SCI in both T3 and T9 injured mice [
14]. Increased CORT is relevant for NE signaling because CORT can increase β2-adrenergic receptor expression and ligand affinity on immune cells [
36,
37]. Therefore, it is possible that CORT could interfere with immune function after acute T9 contusion through enhanced NE sensitivity which decreases CD8 T cell activation.
In this study, we investigated the direct effect of CORT on CD8 T cell proliferation and activation. CORT significantly attenuated proliferation of specific CD8 T cells following activation ex vivo. Furthermore, CORT directly decreased CD8 T cell activation. Effector CD8 T cells treated with CORT had decreased IFNγ production following ex vivo stimulation. Therefore, we believe that systemic changes in immune function and increased stress responses have a negative effect on viral clearance. In addition to increasing NE sensitivity, there are two known mechanisms by which CORT can attenuate the CD8 T cell response. First, CORT has been shown to decrease the ability of antigen-presenting cells to induce proliferation and activation of CD8 T cells [
38,
39]. In addition, CORT may decrease IFNγ production directly in the CD8 T cells as T cells also express glucocorticoid receptors [
40]. Importantly, following in vivo inhibition of CORT, SCI mice had increased expansion of specific CD8 T cells and mice treated with mifepristone had decreased weight loss compared to vehicle-treated mice. Overall, our data indicate that CORT can have a direct effect on activation of specific CD8 T cells.
A recent study has identified adrenal reflex dysfunction as a contributor to immunesuppression following high thoracic injury [
15]. Preventing a spike in CORT following injury while still maintaining normal levels of CORT improved immune cell function and prevented the onset of spontaneous pneumonia [
15]. In this study, however, none of the mice with mid-thoracic injury developed pneumonia, indicating a level-dependent susceptibility to pneumonia infection. In our study, we show impaired CD8 T cell response to viral infection following mid-thoracic injury. Therefore, the innate immune response necessary to clear bacterial infection may be more level-dependent compared to CD8 T cell responses necessary for viral infections, and impaired CD8 T cell function could be a common mechanism of immune dysfunction following both high- and mid-thoracic injuries. It is also possible that mice with mid-thoracic injury would have decreased clearance of pneumonia if infected, as compared to generation of spontaneous infections. Further studies are necessary to identify common mechanisms of immune dysfunction following different levels of SCI and the role of CORT in regulating CD8 T cell function.